CFD modeling of Active Magnetic Regenerator in magnetocaloric refrigeration: a comprehensive review

Prof. Sébastien Poncet
Université de Sherbrooke, Canada

By Prof. Sébastien Poncet, Université de Sherbrooke, Canada.

Magnetocaloric refrigeration is a cooling technology based on the magnetocaloric effect. This technique can be used to attain extremely low temperatures for cryogeny, as well as the ranges encountered in common refrigerators. The US department of energy considered it as one of the top 4 refrigeration technology in terms of technical energy savings potential compared to classical vapor-compression systems. Magnetocaloric refrigerators are composed of a magnet (permanent or electrically powered), a porous regenerator made of a magnetocaloric material, a system that pumps a heat transfer fluid (HTF) and heat exchangers at the thermal sink and source. Most published prototypes try to follow a Brayton like thermodynamic cycle: 1- Fast magnetization of the regenerator; 2- surge of the HTF from the cold to the hot sources; 3- Fast demagnetization of the regenerator and 4- Surge of the HTF from the hot to the cold sources (Kitanovski and Egolf, 2006).
Resolving a large number of thermodynamic cycles to reach steady state and then observe a significative cooling effect remains a very challenging task for numerical methods. It involves indeed the coupling between the fluid flow, the magnetic field and the coupled heat transfer by conduction and forced convection. To do that, numerous numerical models, from 1D to 3D, have been developed so far from the 2000’s. It is straightforward that 1D models are not able to account for the real geometry of packed bed regenerators or of porous regenerators composed of pins. They require also the use of correlations to model the convective heat transfer between the solid material and the fluid as well as the friction at the wall. Regarding the magnetic field, they consider most of the time a uniform magnetic field with an approximated constant demagnetization factor. Some hypotheses may be relaxed when considering 2D models. With the constant increase in computational resources, 3D models have been progressively developed starting from the complete one of Bouchard et al. (2009), who simulated a rather limited porous domain. Recently, Mugica et al. (2018, 2019) proposed a fully coupled 3D solver based only on opensource softwares and libraries. Their results obtained for parallel plate or porous regenerators enable to discuss in details the veracity of all the hypotheses made by 1D and 2D models. It shows also, for the first time, that the laminar fluid flow, the magnetic field and the conjugated heat transfer can be easily decoupled. It opens some interesting perspectives for the development of innovative regenerators for magnetocaloric refrigeration but also for any other caloric technologies. The extension of the solvers to electro-, baro- or elastocaloric based regenerators is indeed relatively straightforward.
Bouchard, J., Nesreddine, H., Galanis, N., 2009. Model of a porous regenerator used for magnetic refrigeration at room temperature. International Journal of Heat and Mass Transfer 52 (5-6), 1223–1229.
Kitanovski, A., Egolf, P. W., 2006. Thermodynamics of magnetic refrigeration. International Journal of Refrigeration 29 (1), 3 – 21.
Mugica, I., Poncet, S., Bouchard, J., 2018. An open source DNS solver for the simulation of active magnetocaloric regenerative cycles. Applied Thermal Engineering 141, 600 – 616.
Mugica I., Poncet S., Bouchard J., 2019. Performance of magnetocaloric pin regenerators by direct numerical simulation, 25th IIR International Congress of Refrigeration, Ed. V. Minea, International Institute of Refrigeration, p.916-923, Montréal, August 24-30, 2019.
Ibai Mugica and Sébastien Poncet
Mechanical Engineering Department, Université de Sherbrooke, Canada


Sébastien Poncet received his engineering diploma and his master degree in physical oceanography from Seatech (Toulon, France) in 2002 and the Ph.D. degree in complex systems from Aix-Marseille University (Marseille, France) in 2005, for a thesis on turbulent rotor-stator interdisk flows (two national awards). He was then assistant lecture for one year before getting a position of assistant professor at Aix-Marseille University in 2006. He joined the University of Sherbrooke as an Associate Professor of Mechanical Engineering in 2014, where he currently is the chairholder of the NSERC/Hydro-Québec/Natural Resources Canada/Emerson Industrial Research Chair in Industrial Energy Efficiency. He is now full professor in the mechanical engineering department and the director of LMFTEUS lab ( His research interests include the experimental characterization of the thermophysical properties of complex heat transfer fluids (nanofluids, slurries, PCMs) and the advanced numerical modelings of thermal systems (supersonic ejector, vortex tube, magnetocaloric refrigeration, heat exchanger, turbomachineries…). He is the coauthor of about 200 research papers whose about 90 in international journals.